Phenom II X4 940 Tested at Stock Speeds

[farlex85]

This test has never been AMD's strong point, never ever has, its like putting a intel up against AMD in the memory test, AMD will always win, so i cant see this benchmark much to go on.

Which tests, he listed three. Actually, i7 easily beats Phenoms in memory benches, triple channel and qpi saw to that. Perhaps AMD can regain some ground when they release the AM3 boards, but I doubt they will reclaim that crown anytime soon.

 

[fullinfusion]

do this x264 HD benchmark

http://www.techarp.com/showarticle.aspx?artno=520

well here im not sure if this is what you wanted

 

[fullinfusion]

LOL, ownage

Dont give up mate

rite back at ya mate

 

[fullinfusion]

well here im not sure if this is what you wanted

why is it showing sb is 700 when its really a 750sb?

 

[HTC]

well here im not sure if this is what you wanted

http://img.techpowerup.org/081223/a11767.jpg

Now: if some owner of a Q6600 would be so kind as to run this bench @ 3 GHz, we could compare results.

 

[Woody112]

That's not how it works. Besides, these haven't been out nearly long enough to determin what kind of voltages they can handle reliably. Generally, when you are on the same basic architecture, die shrinks reduce a chips ability to take voltage.

But you are forgetting that this includes an IMC, whereas the normal Core 2's don't. If you want to compqare voltage differences, you have to compare it to i7.

I under stand that, I was comparing it to the i7-940. Intel rates it at operating from 0.800-1.225v and thats for 2.93ghz.
http://ark.intel.com/cpu.aspx?groupID=37148

I havn't bought into the i-7 yet simply because I'm waiting for the PII to release. If the PII can compete then I'm making the switch to AMD. But if not, then well I'll get an i7.

 

[lemonadesoda]

This test has never been AMD's strong point, never ever has, its like putting a intel up against AMD in the memory test, AMD will always win, so i cant see this benchmark much to go on.

Eh? I quoted three tests. Actually, l only BOTHERED to quote three tests, since the others are essentially similar in what they are doing/showing/calculating, and the point was already demonstrated.

OK, so let's add another: Aquamark3. Q6600 at 2.7 is faster than Phenom2 at 3.0

All that is left to test is wPrime. I havent run that test on my Q6600. Perhaps, when I get the time I will. But really, Q6600 at 2.7 is beating the Phenom2 on the other 4 tests AND it is doing that on POOR DDR1 that we know doesnt have the bandwidth of AMD memory interconnect. Do I really need to run another test?

 

[HTC]

well here im not sure if this is what you wanted

http://img.techpowerup.org/081223/a11767.jpg

Eh? I quoted three tests. Actually, l only BOTHERED to quote three tests, since the others are essentially similar in what they are doing/showing/calculating, and the point was already demonstrated.

OK, so let's add another: Aquamark3. Q6600 at 2.7 is faster than Phenom2 at 3.0

All that is left to test is wPrime. I havent run that test on my Q6600. Perhaps, when I get the time I will. But really, Q6600 at 2.7 is beating the Phenom2 on the other 4 tests AND it is doing that on POOR DDR1 that we know doesnt have the bandwidth of AMD memory interconnect. Do I really need to run another test?

Do you mind running the test of the 1st quote (@ 3 GHz, if possible)?

 

[rodneyhchef]

Here ya go. Looks like AMD does pretty good at this one.

 

[Wile E]

Yeah. AMD is actually quite strong at encoding.

 

[Melvis]

Which tests, he listed three. Actually, i7 easily beats Phenoms in memory benches, triple channel and qpi saw to that. Perhaps AMD can regain some ground when they release the AM3 boards, but I doubt they will reclaim that crown anytime soon.

Pi is the test i meant, i have never seen i7 benchmarks for memory, and i know the AMD X2's will easy beat Phenom X4 in memory tests so if i7 beats Phenom then no surprise there.O well and also that the new i7 is similar to AMD's now anyway with the new OBMC so that would of helped a lot.

I agree i can't seem them reclaiming the crown, but with what ive seen they have improved alot and will equal if not beat C2D and that's pretty good, but we will see at lunch time.
All im interested in is gaming performance, and also encoding , transcoding etc more then anything else.

 

[Melvis]

Eh? I quoted three tests. Actually, l only BOTHERED to quote three tests, since the others are essentially similar in what they are doing/showing/calculating, and the point was already demonstrated.

OK, so let's add another: Aquamark3. Q6600 at 2.7 is faster than Phenom2 at 3.0

All that is left to test is wPrime. I havent run that test on my Q6600. Perhaps, when I get the time I will. But really, Q6600 at 2.7 is beating the Phenom2 on the other 4 tests AND it is doing that on POOR DDR1 that we know doesnt have the bandwidth of AMD memory interconnect. Do I really need to run another test?

Naa its cool, i wasn't specific sorry, i wrote what i meant to farlex85.

I already know about the memory results for both so that's all good.

Thanks for doing all the tests tho

 

[freaksavior]

question?

Does super pi prove speed differences?

if yes then I dont like where they are headed.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
before it gets long gone in this thread

 

[ShadowFold]

AMD has never been good at pi. idk why tho.

 

[Woody112]

Here's a under clocked 9770 for a comparison.

 

[farlex85]

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
before it gets long gone in this thread

Yeah I was trying to find some info on the source code in hopes of perhaps finding out why intel seems to do so much better at this test but I came up empty. I'm imagining there is some instruction set that is perhaps used more heavily in intel architecture. It would be interesting to perhaps try some other software for calculating the same thing as related here. I wonder if the results would be similar. Pifast or Quickpi. Anyone wanna give a go?

Here's some Pifast w/ my e6750 @ 3200. 1.34s using settings up top. Maybe some AMD comparison?

 

[lemonadesoda]

Do you mind running the test of the 1st quote (@ 3 GHz, if possible)?

Sorry, HTC, I cant do 3.0 on my AGP system. I can only OC to 2.7Ghz (300FSB max).

Saw that rodneychef did a Q6600 at 3.0... let me know if a 2.7 DDR1 test is important to you, and I'll do it tomorrow when I'm at that PC

 

[HTC]

Sorry, HTC, I cant do 3.0 on my AGP system. I can only OC to 2.7Ghz (300FSB max).

Saw that rodneychef did a Q6600 at 3.0... let me know if a 2.7 DDR1 test is important to you, and I'll do it tomorrow when I'm at that PC

There's no need: the dude already ran it.

Thanks, anyway!

 

[freaksavior]

Yeah I was trying to find some info on the source code in hopes of perhaps finding out why intel seems to do so much better at this test but I came up empty. I'm imagining there is some instruction set that is perhaps used more heavily in intel architecture. It would be interesting to perhaps try some other software for calculating the same thing as related here. I wonder if the results would be similar. Pifast or Quickpi. Anyone wanna give a go?

Here's some Pifast w/ my e6750 @ 3200. 1.34s using settings up top. Maybe some AMD comparison?

Ok, so amd may have a chance?

 

[farlex85]

Ok, so amd may have a chance?

I don't know it could be that the reason intel does superpi faster is that its just simply that much faster, at least at certain calculations. Or it could be the software, t'would take some investigation methinks.

 

[Melvis]

I don't know it could be that the reason intel does superpi faster is that its just simply that much faster, at least at certain calculations. Or it could be the software, t'would take some investigation methinks.

Id say software or just how intels are designed, because even the old P4's beat AMD back then in Pi. So AMD has no hope in this benchmark.

 

[Darren]

I've heard some ridiculous things in this thread.

They only benchmarked a few applications, Super PI, Cine Bench, Aqua Marks, 3D Marks. These are synthetically benchmarks which do not support the correct number of threads. To make these benchmarks meaningful they at least 10-20 benchmarks, with a full analysis of each test, analysis of results, etc.

People keep talking about Super PI, AMD have never done well in Super PI, so why would they now? Just because a Q6600 performs well in Super PI it doesn't mean its an overall better than the Phenom II 940. This is evident as the Phenom 9950 X4 was just as fast as the Q6600 yet Super PI favoured the Q6600 despite other benchmarks not favouring the Q6600.

We've already seen previews of the Phenom II 940 sometimes beating the i7 920s and trading blows with the i7 940 despite the Phenom using DDR2, it's only slightly slower than the i7 940, think of how the Phenom II would perform with DDR3. Most probably on par or better. So for anyone to suggest that the Q6600 is on par with the Phenom II 940 is just stupid, especially if they are basing it on Super PI.

Let's wait for the proper reviews from established websites before we judge the Phenoms any further.

Edit:

Where did you see those where the 940 trades w/ the i7? I hope your not referring to the Crysis bench of things ending up w/i half a frame of each other. Talk about irrelevant benches.....

I've been following these previews on TPU for a while now, search the forums, they've been posted a few times already.

Half a frame is still trading blows, but Crysis isn't a good benchmark I must admit.

 

[farlex85]

I've heard some ridiculous things in this thread.

They only benchmarked a few applications, Super PI, Cine Bench, Aqua Marks, 3D Marks. These are synthetically benchmarks which do not support the correct number of threads. To make these benchmarks meaningful they at least 10-20 benchmarks, with a full analysis of each test, analysis of results, etc.

People keep talking about Super PI, AMD have never done well in Super PI, so why would they now? Just because a Q6600 performs well in Super PI it doesn't mean its an overall better than the Phenom II 940. This is evident as the Phenom 9950 X4 was just as fast as the Q6600 yet Super PI favoured the Q6600 despite other benchmarks not favouring the Q6600.

We've already seen previews of the Phenom II 940 sometimes beating the i7 920s and trading blows with the i7 940 despite the Phenom using DDR2, it's only slightly slower than the i7 940, think of how the Phenom II would perform with DDR3. Most probably on par or better. So for anyone to suggest that the Q6600 is on par with the Phenom II 940 is just stupid, especially if they are basing it on Super PI.

Let's wait for the proper reviews from established websites before we judge the Phenoms any further.

Where did you see those where the 940 trades w/ the i7? I hope your not referring to the Crysis bench of things ending up w/i half a frame of each other. Talk about irrelevant benches.....

 

[wolf2009]

damn, forget the pi .

x264 encoding is where true CPU performance is at. And AMD looks quite good at beating Intel Penryn Quads at that. Even a 9850 is close enough to them ,

here's Q9450 at 3.2 Ghz

 

[farlex85]

damn, forget the pi .

x264 encoding is where true CPU performance is at. And AMD looks quite good at beating Intel Penryn Quads at that. Even a 9850 is close enough to them ,

here's Q9450 at 3.2 Ghz

Perhaps I'm reading these wrong but they seem to be performing very similarly to each other (penryn and PII) in this bench. Almost identically in fact (unless that closeness is normal).